Small round cell sarcoma tumoroid biobank reveals CIC::DUX4 sarcoma vulnerability to MCL-1 inhibition

Abstract

Small round cell sarcomas (SRCS) are highly aggressive tumors in soft tissues and bone of mostly children and young adults. Despite being different in many aspects, including genetics, possible cell-of-origin, and pathology, patients with any of these entities all receive the same therapeutic regimen. Although several pre-clinical models of Ewing sarcoma have been established, such as cell lines and patient-derived tumor xenografts, few models exist for other SRCS. Here, we describe a pediatric SRCS tumor organoid (tumoroid) biobank containing long-term tumoroid cultures with different translocations, including EWSR1::FLI1, EWSR1::ERG, CIC::DUX4, and BCOR-rearrangements. Using histology, whole genome sequencing and RNA sequencing, we demonstrate that these tumoroids retain histological characteristics, known marker gene expression and chromosomal rearangements of their matching patient tumors. In addition, we compare mutation clusters in the tumoroids across patient-matched longitudinal samples, which shows that cellular heterogeneity is maintained. Drug screening on the tumoroid models unveils entity-specific drug sensitivity to various cytotoxic compounds and targeted compounds, including MCL-1 inhibitors for CIC::DUX4 sarcomas. Taken together, this newly established SRCS patient-derived tumoroid biobank represents a promising source of material for future basic cancer research and drug screening.

Fig. 1. Biobank establishment of small round cell sarcoma cancer tumoroids.

A Flow diagram with an overview of the study design. Tumoroids were established from patient tumor and PDX material and subsequently characterized using whole genome sequencing, RNA sequencing and histology. Drug screens were performed on a subset of ES, CDS and BRS. Lastly, primary tumors and matching tumoroid lines were viably frozen and stored in liquid nitrogen. Created in BioRender. Ringnalda, F. (2025) https://BioRender.com/ovfd63d. B Donut chart (left) representing the different fusions and the number of patients per fusion that were included in the ES, CDS and BRS biobank. Bar chart (right) of the total number of established tumoroid lines from each fusion, subdivided in tumoroids derived from patient tumors and those derived from PDX material. C Representative brightfield microscopy images of EWSR1::FLI1, EWSR1::ERG, EWSR1::FEV, BCOR::CCNB3, KMT2D::BCOR and CIC::DUX4 tumoroids (n = 3, frames of individual area of interest). Scale bar: 200 µm, zoom-in 100 µm. ES Ewing sarcoma, CDS CIC::DUX4 sarcoma, BRS BCOR-rearranged sarcoma. Source data are provided as a Source Data file.

Fig. 2. Histological characterization of small round cell sarcoma tumoroids.

A Representative images of H&E staining of tumor tissue (top) and matching tumoroids (bottom) derived from the indicated fusion-positive tumor types (n = 3, individual tumoroids). Scale bars: 100 µm, zoom-in 50 µm. B Representative images of CD99 staining of tumor tissue (top) and matching tumoroids (bottom) (n = 3, individual tumoroids). Scale bars: 100 µm, zoom-in 50 µm. C Representative images of WT1 staining of tumor tissue (top) and matching tumoroids (bottom) (n = 3, individual tumoroids). Scale bars: 100 µm, zoom-in 50 µm.

Fig. 3. Genetic characterization of small round cell sarcoma tumoroid models.

A Oncoplot depicting somatic mutations, copy number alterations (CNAs) and gene fusions identified in ES, CDS and BRS tumors and matching tumoroids. For samples in which the germline sample was not available (*), mutations in known tumor driver genes are indicated. For samples for which no sequencing data was available, the mutational information provided by the ITCC-P4 consortium is indicated (+). B Circos plots of ES-010 tumor and tumoroid. The circos plots show, from outside inwards: chromosomes, single-nucleotide variants (SNVs) (dark blue), copy number alterations (CNAs) as gained (red) or lost (light blue) regions and translocations (red line). C Circos plots of tumoroids derived from one EWSR1::FLI1 patient (top row; diagnostic [ES-016], first relapse [ES-055] and second relapse [ES-080] samples) and one CIC::DUX4 patient (bottom row; diagnostic [CDS-090], resection [CDS-032] and relapse [CDS-045] samples) at different timepoints during disease progression. The circos plots show, from outside inwards: chromosomes, single-nucleotide variants (SNVs) (dark blue), copy number alterations (CNAs) as gained (red) or lost (light blue) regions and translocations (red line). D River plot depicting differential clusters of mutations between time of diagnosis (ES-016), first relapse (ES-055) and second relapse (ES-080) tumor sample from one EWSR1::FLI1 patient. E The mean variant allele frequency of individual clusters of mutations in tumor and matching tumoroid samples in diagnostic (ES-016), first relapse (ES-055) and second relapse (ES-080) samples from one EWSR1::FLI1 patient. F Contribution of somatic mutational signatures per tumor and matching tumoroid model. SBS single base substitution, TT Tumor Tissue, TO TumorOid, ES Ewing sarcoma, CDS CIC::DUX4 sarcoma, BRS BCOR-rearranged sarcoma. Source data are provided as a Source Data file.

Fig. 4. Transcriptomic profiling of small round cell sarcoma tumoroid models.

A Principal component analysis (PCA) of top 5000 most variable genes, dots are colored based on tumor entity and shaped for tumoroid, tissue or PDX tissue. B Heatmap showing marker gene expression of the different tumor entities in tumor and tumoroid samples. C Heatmap showing the expression of the top 20 differentially expressed genes per tumor entity that have a CHIPseq binding peak of their characteristic (fusion-)protein in close proximity to the transcription start site. D Dot plot showing row-normalized mean expression values of genes that are targeted by our in-house drug library in the different tumor entities. Source data are provided as a Source Data file.

Fig. 5. Drug screening on small round cell sarcoma tumoroids.

A Schematic overview of the tumoroid drug screening workflow. Created in BioRender. Ringnalda, F. (2025) https://BioRender.com/mnd48e5. B Clustered heatmap of AUC values of each tumoroid model (x axis) and compound (y axis), showing AUC values after 120 h drug treatment in a dose-range of 0.1 nM to 10 µM. Low AUC (red) indicates high drug sensitivity while high AUC (blue) indicates low drug sensitivity. Highlighted clusters include (1) chemotherapy, proteases and histone deacetylase inhibitors, (2) mTOR inhibitors, (3) E3 ubiquitin-protein ligases and (4) MCL-1 inhibitors. C Clustered heatmap showing AUC values after chemotherapy treatment, compounds that are routinely used in ES and ELS patients marked in red. D Box plots showing the mean AUCs values of mTOR inhibitors in the indicated ES (n = 13, tumoroid lines), CDS (n = 6, tumoroid lines) and BRS (n = 2, tumoroid lines) tumoroids entities. The boxes represent the interquartile range, the line inside the box marks the median, and the whiskers extend to the lowest and highest values. Error bars represent SEM of individual indicated ES, CDS and BRS tumoroid enitities. P-values were calculated using a one-way Anova, Dunnet multiple comparisons test:*<0.05, **<0.005, ns not significant. Exact adjusted P-values: AZD8055, EWSR1::FLI1 vs. EWSR1::ERG/FEV = 0.0033, EWSR1::FLI1 vs. BCOR::CCNB3/KMT2D = 0.0157, EWSR1::FLI1 vs. CIC::DUX4 = 0.0035. Sirolimus, EWSR1::FLI1 vs. EWSR1::ERG/FEV = 0.0147, EWSR1::FLI1 vs. BCOR::CCNB3/KMT2D = ns, EWSR1::FLI1 vs. CIC::DUX4 = 0.0281. Temsirolimus, EWSR1::FLI1 vs. EWSR1::ERG/FEV = ns, EWSR1::FLI1 vs. BCOR::CCNB3/KMT2D = ns, EWSR1::FLI1 vs. CIC::DUX4 = ns. E Box plots showing the mean AUCs values of P53 pathway activators in the indicated ES, CDS and BRS tumoroids entities. The boxes represent the interquartile range, the line inside the box marks the median, and the whiskers extend to the lowest and highest values. Error bars represent SEM of individual indicated TP53 wildtype (p53^WT) (n = 14) vs TP53 mutant (p53^MUT) (n = 7) tumoroids lines. P-values were calculated using a two-tailed Student’s t test: ****<0.0001. Exact P-values: Idasanutlin ≤ 0.0001, Milademetan ≤ 0.0001, Siramadlin ≤ 0.0001. F Clustered heatmap of compounds targeting the apoptosis pathway, CDS tumoroids showing specific sensitivity towards MCL-1 inhibitors (in red). G Validation screen with MCL-1 inhibitors, dose-response curves of S63845, MIK665 and AZD5991 for indicated ES and CDS tumoroid cultures. Error bars represent SD, each consisting of triplicate measurements. SEM standard error of mean, SD standard deviation, ES Ewing sarcoma, CDS CIC::DUX4 sarcoma, BRS BCOR-rearranged sarcoma. Source data are provided as a Source Data file.